04.04.2013 Views

View - American Journal of Botany

View - American Journal of Botany

View - American Journal of Botany

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>American</strong> <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 86(9): 1290–1300. 1999.<br />

PHYLOGENETIC ANALYSES OF ANDROMEDEAE<br />

(ERICACEAE SUBFAM. VACCINIOIDEAE) 1<br />

KATHLEEN A. KRON, 2,3 WALTER S. JUDD, 4 AND DARREN M. CRAYN 5<br />

3 Department <strong>of</strong> Biology, Wake Forest University, Winston-Salem, North Carolina 27109-7325;<br />

4 Department <strong>of</strong> <strong>Botany</strong>, University <strong>of</strong> Florida, Gainesville, Florida 32611-8526; and<br />

5 School <strong>of</strong> Biological Science, University <strong>of</strong> New South Wales, Sydney 2052, Australia<br />

Phylogenetic relationships within the Andromedeae and closely related taxa were investigated by means <strong>of</strong> cladistic<br />

analyses based on phenotypic (morphology, anatomy, chromosome number, and secondary chemistry) and molecular (rbcL<br />

and matK nucleotide sequences) characters. An analysis based on combined molecular and phenotypic characters indicates<br />

that the tribe is composed <strong>of</strong> two major clades—the Gaultheria group (incl. Andromeda, Chamaedaphne, Diplycosia, Gaultheria,<br />

Leucothoë, Pernettya, Tepuia, and Zenobia) and the Lyonia group (incl. Agarista, Craibiodendron, Lyonia, and<br />

Pieris). Andromedeae are shown to be paraphyletic in all analyses because the Vaccinieae link with some or all <strong>of</strong> the<br />

genera <strong>of</strong> the Gaultheria group. Oxydendrum is sister to the clade containing the Vaccinieae, Gaultheria group, and Lyonia<br />

group. The monophyly <strong>of</strong> Agarista, Lyonia, Pieris, and Gaultheria (incl. Pernettya) is supported, while that <strong>of</strong> Leucothoë<br />

is problematic. The close relationship <strong>of</strong> Andromeda and Zenobia is novel and was strongly supported in the molecular (but<br />

not morphological) analyses. Diplycosia, Tepuia, Gaultheria, and Pernettya form a well-supported clade, which can be<br />

diagnosed by the presence <strong>of</strong> fleshy calyx lobes and methyl salicylate. Recognition <strong>of</strong> Andromedeae is not reflective <strong>of</strong> our<br />

understanding <strong>of</strong> geneological relationships and should be abandoned; the Lyonia group is formally recognized at the tribal<br />

level.<br />

Key words: Andromedeae; Ericaceae; Lyonieae; matK; molecular systematics; rbcL.<br />

Andromedeae (13 genera, subfam. Vaccinioideae; see<br />

Stevens 1971, 1995) are a heterogeneous group <strong>of</strong> doubtful<br />

monophyly (Judd and Kron, 1993; Kron and Chase,<br />

1993; Stevens, 1995; Kron, 1996, 1997, and unpublished<br />

data; Kron and Judd, 1997; Kron et al., 1998). Members<br />

<strong>of</strong> this ‘‘group’’ are characterized by alternate leaves with<br />

well-developed blades, usually axillary inflorescences,<br />

sympetalous, urceolate flowers with <strong>of</strong>ten appendaged<br />

anthers, superior ovaries with numerous ovules, and usually<br />

loculicidal capsules. The members <strong>of</strong> this tribe are<br />

distinguished from the phenetically similar Vaccinieae by<br />

their superior (vs. inferior) ovaries and usually dry, capsular<br />

(vs. fleshy, baccate) fruits (Stevens, 1969, 1971,<br />

1995; Anderberg, 1993; Judd and Kron, 1993). They are<br />

easily distinguished from members <strong>of</strong> the Arbuteae,<br />

which also have urceolate corollas and superior ovaries,<br />

by their well-developed, lignified, fiber sheath surrounding<br />

the midrib in the leaves (vs. such fibers poorly developed<br />

or lacking), anthers that become inverted early<br />

in development (vs. inverting only very late), and with<br />

numerous (vs. few or only one) ovules per locule, dry<br />

fruits (vs. fleshy fruits with a prominent layer <strong>of</strong> fibers<br />

around inside <strong>of</strong> pericarp), and the lack (vs. presence) <strong>of</strong><br />

high concentrations <strong>of</strong> ellagic acid (Matthews and Knox,<br />

1 Manuscript received 6 July 1998; revision accepted 23 February<br />

1999.<br />

This study was supported by National Science Foundation grant<br />

DEB-9407350 (KAK). Oligonucleotide synthesis was performed in the<br />

DNA Synthesis Core Laboratory <strong>of</strong> the Cancer Center <strong>of</strong> Wake Forest<br />

University supported in part by NIH grant CA-12197. The authors thank<br />

the following people and institutions for sending herbarium specimens<br />

and providing material for DNA extraction—Roger Hyam, Royal Botanic<br />

Garden, Edinburgh, Royal Botanic Garden, Kew, and New York<br />

Botanical Garden, Kent Perkins for assistance in processing specimens;<br />

and Norris Williams for making available certain computing facilities.<br />

2 Author for correspondence.<br />

1290<br />

1926; Stevens, 1969, 1971; Anderberg, 1993; Judd and<br />

Kron, 1993).<br />

Most genera <strong>of</strong> Andromedeae have been considered in<br />

one <strong>of</strong> two informal subgroups, i.e., the Lyonia group,<br />

comprising Lyonia Nutt. (36 spp.), Craibiodendron W.<br />

W. Smith (five spp.), Pieris D. Don (incl. Arcterica Coville)<br />

(seven spp.), and Agarista (31 spp., incl. Agauria<br />

J. D. Hooker) and the Gaultheria group, including Leucothoë<br />

D. Don (eight spp.), Zenobia D. Don (one sp.),<br />

Gaultheria L. (115 spp.), Pernettya Gaud. (14 spp., but<br />

this genus is <strong>of</strong>ten included within Gaultheria; see Middleton<br />

and Wilcock, 1990a; Middleton, 1991b), Tepuia<br />

Camp (seven spp.), and Diplycosia Blume (97 spp., incl.<br />

Pernettyopsis King and Gamble; see Stevens, 1995) (Stevens,<br />

1970a, 1971, 1995; Judd, 1979; Middleton, 1991a,<br />

b; Kron and Judd, 1997). Andromeda L. (two spp.), Chamaedaphne<br />

Moench. (one sp.), and Oxydendrum DC (one<br />

sp.). have been considered to be taxonomically isolated<br />

(Stevens, 1969, 1971). The Lyonia and Gaultheria groups<br />

are somewhat easier to characterize than the tribe as a<br />

whole. The members <strong>of</strong> the Lyonia group usually have<br />

multicellular hairs with biseriate stalks, slender, geniculate<br />

filaments, and short and rather broad anthers with<br />

white disintegration tissue at the anther–filament junction.<br />

Staminal appendages, when present, are spurs borne<br />

either on the filaments or dorsally on the anthers; the testa<br />

cells are much elongated and have thin walls; the foliar<br />

stomata are usually anomocytic; the upper epidermis <strong>of</strong><br />

the leaf is <strong>of</strong>ten lignified; bands <strong>of</strong> fibers are found in the<br />

secondary phloem; and chromosome numbers are all x <br />

12. In contrast, the members <strong>of</strong> the Gaultheria group can<br />

be characterized by their multicellular hairs with multiseriate<br />

stalks; their stouter, straight filaments; <strong>of</strong>ten longer<br />

anthers with disintegration tissue on the anthers and terminal<br />

awns or lacking both awns and disintegration tis-


September 1999] KRON ET AL.—PHYLOGENETICS OF ANDROMEDEAE<br />

1291<br />

sue; testa cells variable in shape, but <strong>of</strong>ten about as broad<br />

as long and distinctly thickened; <strong>of</strong>ten paracytic stomata;<br />

unlignified cells in the phloem not occurring in bands;<br />

and chromosome base numbers <strong>of</strong> x 11, 12, 13, and<br />

18.<br />

Monophyly <strong>of</strong> the Andromedeae has never been rigorously<br />

assessed. Therefore, in this paper, we focus on<br />

the question <strong>of</strong> the group’s monophyly and its appropriate<br />

circumscription. In addition, we investigate here the<br />

monophyly <strong>of</strong> both the Lyonia group and the Gaultheria<br />

group. The monophyly <strong>of</strong> the Gaultheria group has never<br />

been investigated, while the monophyly <strong>of</strong> the Lyonia<br />

group has been considered only in a preliminary fashion<br />

(Judd, 1979; Kron and Judd, 1997). Selected members <strong>of</strong><br />

the Vaccinieae and Epacridoideae are included in these<br />

cladistic analyses because <strong>of</strong> previous analyses that indicated<br />

a close relationship to members <strong>of</strong> the Andromedeae<br />

(Anderberg, 1993; Kron and Chase, 1993; Kron,<br />

1996, 1997; Kron et al., 1998).<br />

Ericaceae show a great deal <strong>of</strong> morphological variability<br />

and have been well studied from the standpoint <strong>of</strong><br />

morphology, anatomy, embryology, and secondary chemistry<br />

(Niedenzu, 1890; Artopoeus, 1903; Samuelsson,<br />

1913; Matthews and Knox, 1926; Cox, 1948; Palser,<br />

1951, 1952, 1954, 1958, 1961a, b; Chou, 1952; Copeland,<br />

1954; Safijowska, 1960; Paterson, 1961; Wood,<br />

1961; Lems, 1962, 1964; Watson, 1962, 1965; Leins,<br />

1964; Towers, Tse, and Maas, 1966; Harborne, 1969; Stevens,<br />

1969, 1970a, b, 1971; Harborne and Williams,<br />

1973; Villamil and Palser, 1981; Vander Kloet, 1983,<br />

1988; Baas, 1985; Rao and Chakraborti, 1985; Middleton<br />

and Wilcock, 1990a, b; Middleton, 1991a; Odell and<br />

Vander Kloet, 1991; Anderberg, 1994; Powell et al.,<br />

1997). In addition, many <strong>of</strong> the genera considered here<br />

have been recently monographed or are part <strong>of</strong> recent<br />

floristic treatments (Sleumer, 1959, 1966, 1967; Hersey<br />

and Vander Kloet, 1976; Dorr, 1980; Melvin, 1980; Judd,<br />

1981, 1982, 1984, 1986, 1990, 1995a, b, c; Judd and<br />

Hermann, 1990; Luteyn, 1991, 1995a, b, c, 1996; Middleton,<br />

1991b; Luteyn et al., 1996), and their general phylogenetic<br />

placement within the Ericaceae can be assessed<br />

by reference to several higher level phylogenetic analyses<br />

that use morphology (Anderberg, 1993; Judd and Kron,<br />

1993; Kron and Judd, 1997) and rbcL, matK, and nr18s<br />

sequence data (Kron and Chase, 1993; Kron, 1996, 1997,<br />

and unpublished data; Kron and King, 1996; Kron and<br />

Judd, 1997; Kron et al., 1999).<br />

Current studies by Kron (Kron, 1997; Kron and Judd,<br />

1997, and work in progress) indicate that matK is an appropriate<br />

gene to use for phylogenetic analysis in Ericaceae<br />

s.l. (sensu lato). Our knowledge <strong>of</strong> morphological<br />

and molecular variation within the Andromedeae certainly<br />

is now sufficient to undertake phylogenetic analyses<br />

<strong>of</strong> this group, and numerous recent studies have demonstrated<br />

the effectiveness <strong>of</strong> combined analyses <strong>of</strong> phenotypic<br />

and molecular data.<br />

MATERIALS AND METHODS<br />

Representatives <strong>of</strong> all 13 genera <strong>of</strong> Andromedeae were sampled for<br />

this study. Twenty-eight taxa were sampled for the phenotypic cladistic<br />

analyses, while 29 taxa were included in the molecular analyses (Table<br />

1). This study specifically tested the monophyly <strong>of</strong> the tribe Andro-<br />

TABLE 1. Ericalean species included in cladistic analyses <strong>of</strong> the Andromedeae.<br />

All species were used in both morphological and molecular<br />

analyses except Harrimanella hypnoides (see text); Gen-<br />

Bank accession numbers are in brackets. a All species were traditionally<br />

placed in Andromedeae unless otherwise indicated.<br />

Enkianthus campanulatus (Miq.) Nichols—outgroup [matK, GBAN-<br />

U61344; rbcL, GBAN-L12616]<br />

Harrimanella hypnoides (L.) Cov.—outgroup [matK, GBAN-U61315:<br />

rbcL, GBAN-U82766]<br />

Sprengelia incarnata Smith—as representative the epacrid clade, and<br />

as outgroup [matK, GBAN-AF015645; rbcL, GBAN-80421]<br />

Prionotes cerinthoides R. Br.—as representative <strong>of</strong> the epacrid clade,<br />

and as outgroup [matK, GBAN-AF015642; rbcL, GBAN-U79743]<br />

Agarista populifolia (Lam.) Judd [matK, GBAN-U61306; rbcL,<br />

GBAN-AF124589]<br />

Agarista salicifolia (Comm. ex Lam.) G. Don [matK, GBAN-U61313;<br />

rbcL, GBAN-AF124588]<br />

Andromeda polifolia L. [matK, GBAN-AF124569; rbcL, GBAN-<br />

AF124572]<br />

Chamaedaphne calyculata (L.) Moench. [matK, GBAN-AF015630;<br />

rbcL, GBAN-L12606]<br />

Craibiodendron yunnanense W. W. Smith [matK, GBAN-U61307;<br />

rbcL, GBAN-AF124589]<br />

Diplycosia acuminata Becc. [matK, GBAN-AF124563; rbcL, GBAN-<br />

AF124586]<br />

Gaultheria miqueliana Takeda [matK, GBAN-AF124567; rbcL,<br />

GBAN-AF124590]<br />

Gaultheria shallon Pursh [matK, GBAN-AF124565; rbcL, GBAN-<br />

AF124574]<br />

Pernettya tasmanica J.D. Hooker [matK, GBAN-AF124568; rbcL,<br />

GBAN-U82765]<br />

Leucothoë racemosa (L.) Gray [matK, GBAN-AF124564; rbcL,<br />

GBAN-U83915]<br />

Leucothoë fontanesiana (Steudel) Sleumer [matK, GBAN-AF124570;<br />

rbcL, GBAN-AF124585]<br />

Lyonia ligustrina (L.) DC. [matK, GBAN-U61311; rbcL, GBAN-<br />

AF124578]<br />

Lyonia ferruginea (Walter) Nutt. [matK, GBAN-U61312; rbcL,<br />

GBAN-AF124584]<br />

Lyonia lucida (Lam.) K. Koch [matK, GBAN-U61308; rbcL, GBAN-<br />

U82764]<br />

Lyonia ovalifolia (Wallich) Drude ([matK, GBAN-U61305; rbcL,<br />

GBAN-AF124580]<br />

Oxydendrum arboreum (L.) DC. [matK, GBAN-AF124562; rbcL,<br />

GBAN-AF124583]<br />

Pieris phillyreifolia (W.J. Hooker) DC. [matK, GBAN-U61309; rbcL,<br />

GBAN-AF124573]<br />

Pieris floribunda (Pursh) Bentham and Hooker [matK, GBAN-<br />

U61304; rbcL, GBAN-AF124577]<br />

Pieris formosa (Wallich) D. Don [matK, GBAN-U61303; rbcL,<br />

GBAN-AF124581]<br />

Pieris nana (Maxim.) Makino [matK, GBAN-U61310; rbcL, GBAN-<br />

AF124582]<br />

Satyria warszewiczii Klotzsch [matK, GBAN-U61314; rbcL, GBAN-<br />

AF124579]<br />

Tepuia cardonae A. C. Smith [matK, GBAN-AF124566; rbcL,<br />

GBAN-AF124575]<br />

Vaccinium macrocarpon Aiton [matK, GBAN-U61316; rbcL, GBAN-<br />

L12625]<br />

Vaccinium meridionale Sw. [matK, GBAN-U89759; rbcL, GBAN-<br />

AF124576]<br />

Zenobia pulverulenta (Bartram ex Willd.) Pollard [matK, GBAN-<br />

AF124571; rbcL, GBAN-L12626]<br />

a The prefix GBAN- has been added for linking the online version <strong>of</strong><br />

<strong>American</strong> <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> to GenBank but is not part <strong>of</strong> the actual<br />

GenBank accession number.


1292 AMERICAN JOURNAL OF BOTANY<br />

[Vol. 86<br />

medeae and thus was not primarily concerned with generic limits. However,<br />

most <strong>of</strong> the genera within the tribe have been monographed and<br />

their monophyly is supported by morphological synapomorphies (Judd<br />

1981, 1982, 1984, 1986, 1995b). As discussed in Kron and Judd (1997;<br />

see also Weins, 1998) supraspecific taxa (i.e., genera and sections) were<br />

represented by particular species.<br />

Character scorings for most <strong>of</strong> the 60 phenotypic characters (i.e.,<br />

morphological, anatomical, and embryological—for convenience hereafter<br />

merely referred to as morphological characters) (Table 2) are based<br />

on the authors’ observations <strong>of</strong> herbarium material and, where possible,<br />

living material, supplemented by revisionary and monographic studies<br />

(see references cited above). Character scorings relating to chromosome<br />

number and chemical features mainly were taken from the literature<br />

(see Table 3).<br />

Most morphological characters were readily divisible into discrete<br />

states, thus avoiding arbitrary decisions relating to state delimitation<br />

(Stevens, 1991). A few problems are noted in Table 3. Some characters<br />

could not be included in the analyses because they showed too much<br />

infrataxon variation or could not be delimited into discrete states, e.g.,<br />

leaf size, inflorescence structure, bracteole position, calyx size, corolla<br />

shape and size, fruit shape, and placenta position. All multistate characters<br />

were considered to be unordered (Table 2). Species varying in<br />

particular characters were scored as ‘‘?’’ (and indicated as ‘‘variable’’<br />

in Table 3).<br />

Total DNA was extracted from fresh or silica-gel dried (Chase and<br />

Hills, 1991) leaves using the modified CTAB (hexadecyltrimethylammonium<br />

bromide) procedure <strong>of</strong> Doyle and Doyle (1987). The matK<br />

gene <strong>of</strong> the chloroplast DNA was amplified by the polymerase chain<br />

reaction (PCR) via the methods described in Olmstead et al. (1992)<br />

using the same parameters as described in Kron et al. (1999). Primer<br />

sequences from Johnson and Soltis (1994, 1995) and Steele and Vilgalys<br />

(1994) were used for matK PCR and sequencing. For the rbcL<br />

gene, primers and PCR protocols followed those <strong>of</strong> Olmstead et al.<br />

(1992). Amplified products were cleaned using Microcon 100 filter<br />

tubes (Fisher Scientific, Atlanta, Georgia). The nucleotide sequencing<br />

was performed at the Wake Forest School <strong>of</strong> Medicine, DNA Core<br />

Laboratory on an ABI 377 Automated DNA Sequencer (Perkin-Elmer,<br />

Foster City, California). Raw sequences (for both rbcL and matK) were<br />

edited using Sequencher 3.0 (Gene Codes Corp., Ann Arbor, Michigan).<br />

The edited sequences were visually aligned.<br />

Voucher information for all taxa sampled in this study can be obtained<br />

from KAK. All sequences can be obtained through GenBank<br />

(Table 1).<br />

Phylogenetic analyses—Analyses <strong>of</strong> morphological data included<br />

representatives <strong>of</strong> outgroups identified as appropriate based on previous<br />

studies (Judd and Kron, 1993; Kron and Chase, 1993; Kron, 1996,<br />

1997; Kron and Judd, 1997). Enkianthus campanulatus, Sprengelia incarnata,<br />

and Prionotes cerinthoides were used as outgroups. [Harrimanella<br />

(used in the molecular analyses) was not used as an outgroup<br />

taxon because its very different morphology makes determining character<br />

homology difficult.] Analyses <strong>of</strong> the morphological data were<br />

rooted with Enkianthus, allowing the remaining outgroup taxa to resolve<br />

simultaneously with the ingroup taxa. The analyses employed the<br />

branch-and-bound algorithm (ie-) with extended branch swapping (bb*)<br />

<strong>of</strong> the Hennig86, version 1.5, computer s<strong>of</strong>tware developed by Farris<br />

(1988), and the branch-and-bound option <strong>of</strong> PAUP (Phylogenetic Analysis<br />

Using Parsimony) 3.1.1. (Sw<strong>of</strong>ford, 1993).<br />

As in the morphological study, analysis <strong>of</strong> molecular data (rbcL and<br />

matK nucleotide sequences) included representatives <strong>of</strong> outgroups identified<br />

as appropriate based on previous studies (Judd and Kron, 1993;<br />

Kron and Chase, 1993) and also on the basis <strong>of</strong> preliminary analysis <strong>of</strong><br />

larger matK and rbcL data sets (Kron et al., 1999). These indicate appropriate<br />

potential outgroups as Enkianthus campanulatus, Harrimanella<br />

hypnoides, Sprengelia incarnata, and Prionotes cerinthoides; the<br />

latter two represent the epacrid clade (see Kron, 1997). Analyses <strong>of</strong> the<br />

molecular data were run with Enkianthus designated as the outgroup,<br />

allowing the remaining outgroup taxa to resolve simultaneously with<br />

the ingroup taxa. The few indels that occurred were treated as missing<br />

data. All characters were weighted equally in all analyses. The analyses<br />

were run using the heuristic search option (MULPARS, TBR, 1000<br />

random replicates) <strong>of</strong> PAUP 3.1.1 (Sw<strong>of</strong>ford, 1993). Three measures <strong>of</strong><br />

internal support were used to assess clade robustness when appropriate:<br />

parsimony jackknife (Farris et al., 1996) bootstrap (100 random replicates,<br />

both analyses) and decay (Autodecay 3.0; Mishler, Donoghue,<br />

and Albert, 1991; Eriksson and Wikstrom, 1995).<br />

Combined parsimony analyses (1000 random replicates, MULPARS,<br />

TBR) based on rbcL matK nucleotide sequences, and rbcL matK<br />

morphological data were performed using Enkianthus campanulatus,<br />

Prionotes cerinthoides, and Sprengelia incarnata, as outgroup taxa (and<br />

analysed by PAUP 3.1.1). Bootstrap (100 replicates) and decay analyses<br />

were performed to assess internal support for relationships obtained in<br />

the heuristic analyses.<br />

RESULTS<br />

Morphological data—The morphological analyses<br />

(Hennig86 and PAUP 3.1.1) resulted in the generation <strong>of</strong><br />

36 equally parsimonious trees <strong>of</strong> 139 steps, a consistency<br />

index (CI) <strong>of</strong> 0.48, and a retention index (RI) <strong>of</strong> 0.68.<br />

The topologies represented in these trees, however, are<br />

all quite similar, differing mainly in the placement <strong>of</strong> Andromeda<br />

and a few members <strong>of</strong> the Gaultheria group.<br />

The strict consensus (Fig. 1) and a representative cladogram<br />

(Fig. 2) are presented.<br />

Oxydendrum consistently is the sister group to all other<br />

ingroup taxa, which are are united on the basis <strong>of</strong> their<br />

inflorescences developing on shoots <strong>of</strong> the previous season<br />

(character no. 22) and the pedicel articulated with the<br />

flower (no. 26). Oxydendrum is quite distinctive because<br />

<strong>of</strong> its Calluna-type pith (no. 2-2), deciduous leaves (no.<br />

10-0), and flowers in which all traces to the floral organs<br />

leave the elongated floral axis separately (Palser, 1952).<br />

The monophyly <strong>of</strong> Vaccinieae (as represented by Vaccinium<br />

macrocarpon, V. meridionale, and Satyria warszewiczii)<br />

is supported by their inferior ovary (no. 46)<br />

and indehiscent fruit (no. 51-2), and this group is represented<br />

in most trees. Fleshy fruits (no. 49) are also synapomorphic<br />

for these species under a DELTRAN (see<br />

Wiley et al. [1991] for discussion <strong>of</strong> ACCTRAN and<br />

DELTRAN) optimization. However, Tepuia sometimes is<br />

placed within this clade, while in other trees it links with<br />

Diplycosia because they share apical connate bracteoles<br />

(no. 25) and methyl salicylate (no. 58). The Diplycosia<br />

Tepuia clade, when present, may be sister to Vaccinieae<br />

(Fig. 2), due, in part, to their anther tubules (no.<br />

41), or placed elsewhere in the tree, but these species are<br />

always phylogenetically adjacent to members <strong>of</strong> Vaccinieae.<br />

Thus, in the strict consensus tree (Fig. 1) the<br />

monophyly <strong>of</strong> Vaccinieae is not evident.<br />

Several genera, i.e., Pieris, Lyonia, Agarista, and<br />

Gaultheria s.l., are consistently indicated as monophyletic.<br />

The monophyly <strong>of</strong> Pieris is supported by the inflorescence<br />

exposed during the winter (no. 23) and valvate<br />

calyx lobes (no. 27). Paired appendages located at the<br />

anther–filament junction (no. 39-1) are also synapomorphic<br />

under a DELTRAN optimization. Synapomorphies<br />

for the species <strong>of</strong> Lyonia include the corolla with<br />

multicellular hairs (no. 31), disintegration tissue on the<br />

staminal appendages (no. 43), ovary with multicellular


September 1999] KRON ET AL.—PHYLOGENETICS OF ANDROMEDEAE<br />

1293<br />

TABLE 2. Characters used in morphology-based cladistic analysis <strong>of</strong> Andromedeae. Plesiomorphic characters are listed first, followed by apomorphic<br />

ones. (Polarity decisions based on trees rooted with Enkianthus.)<br />

1. Secondary phloem without bands <strong>of</strong> fibers (0); secondary phloem with bands <strong>of</strong> fibers (1).<br />

2. Pith heterogeneous (0); pith homogeneous (1); pith Calluna-type (2). [unordered]<br />

3. Bud scales numerous to 2, if the latter then minute (0); bud scales 2, enlarged (1).<br />

4. Leaves convolute in bud (0); leaves revolute in bud (1).<br />

5. Leaves alternate (0); leaves whorled (1).<br />

6. Leaf venation pinnate (0); leaf venation parallel (1).<br />

7. Vein reticulum open, the secondary and tertiary veins not equally prominent (0); vein reticulum dense, the secondary and tertiary veins <br />

equally prominent (1).<br />

8. Leaf margin serrate or serrulate (0); leaf margin entire (1).<br />

9. Leaf margin lacking raised glands (0); leaf margin with raised glands (1).<br />

10. Leaves deciduous (0); leaves persistent (1). [polarity questionable; see Table 3]<br />

11. Leaf midrib (as seen in cross section) unifacial (0); leaf midrib bifacial (1).<br />

12. Fiber strands absent from leaf mesophyll (0); fiber strands present in leaf mesophyll (1).<br />

13. Leaf epidermis not lignified (0); leaf epidermis at least slightly lignified (1).<br />

14. Leaf epidermal cells not elongated (as seen in cross section) (0); epidermal cells elongated (1).<br />

15. Hypodermis lacking (0); hypodermis present (1).<br />

16. Multicellular hairs on leaves present (0); multicellular hairs lacking (1).<br />

17. Multicellular hairs with small glandular head (or hairs not glandular) (0); multicellular hairs with expanded and elongated, glandular head (1).<br />

18. Peltate scales lacking (0); peltate scales present (with margin irregular-erose) (1); peltate scales present (with margin entire) (2). [unordered]<br />

19. Stalk <strong>of</strong> multicellular hair multiseriate (0); stalk <strong>of</strong> multicellular hair biseriate (1).<br />

20. Stomata anomocytic (0); stomata paracytic (1); stomata tetracytic (2). [unordered]<br />

21. Inflorescences monotelic, i.e., terminal flowers present (0); inflorescences polytelic, i.e., terminal flowers lacking (1).<br />

22. Inflorescences developing (and producing flowers and fruits) on shoots <strong>of</strong> the current season (0); inflorescences developing (and producing<br />

flowers and fruits) on shoots <strong>of</strong> the previous season (1).<br />

23. Inflorescences within bud (0); inflorescences exposed during winter or dormant season (1).<br />

24. Pedicel lacking bracteoles (0); pedicel with two bracteoles (1); pedicel with several bracteoles (2). [unordered]<br />

25. Bracteoles distinct, variably positioned (0); bracteoles connate, apical (1).<br />

26. Pedicel not articulated with flower (0); pedicel articulated with flower (1).<br />

27. Calyx <strong>of</strong> imbricate lobes (0); calyx <strong>of</strong> valvate lobes (1).<br />

28. Calyx lobes brown, dry at maturity (0); calyx lobes usually colorful, fleshy at maturity (1).<br />

29. Adaxial calyx stomata lacking (0); adaxial calyx stomata present (1).<br />

30. Corolla lacking unicellular hairs on inside surface (0); corolla with unicellular hairs on inside surface (1).<br />

31. Corolla lacking multicellular hairs on outer surface (0); corolla with multicellular hairs on outer surface (1).<br />

32. Abaxial corolla stomata present (0); abaxial corolla stomata absent (1).<br />

33. Filaments straight (0); filaments S-shaped (1).<br />

34. Filaments distinct (0); filaments connate (1).<br />

35. Filaments free from corolla (0); filaments adnate to corolla (1).<br />

36. Filaments with unicellular hairs (0); filaments papillose or minutely roughened (1); filaments smooth (2). [unordered]<br />

37. Anthers inverting late in development (0); anthers inverting early in development (1).<br />

38. Anthers with projections (awns or spurs) (0); such projections absent (1).<br />

39. Projections borne on anther, erect to deflexed (0); projections borne at junction <strong>of</strong> anther and filament (1); projections borne on filament (2).<br />

[unordered]<br />

40. Projections single (or lacking) (0); projections divided (1).<br />

41. Tubules on anthers absent (0); tubules on anthers present (1).<br />

42. Disintegration tissue on back <strong>of</strong> anthers absent (0); disintegration tissue present (1).<br />

43. Disintegration tissue not extending onto the spurs (or spurs lacking) (0); disintegration tissue extending onto the spurs (1).<br />

44. Anthers open by longitudinal slits (0); anthers open by ‘‘terminal’’ pores (1).<br />

45. Pollen shed as monads (0); pollen shed as tetrads (or modified tetrads) (1).<br />

46. Ovary superior (0); ovary inferior (1).<br />

47. Ovary lacking multicellular gland-headed hairs (0); ovary with multicellular gland-headed hairs (or homologous hair types, e.g., scales) (1).<br />

48. Locules not divided (0); locules incompletely divided by outgrowth <strong>of</strong> ovary wall (1).<br />

49. Fruit dehiscent (0); fruit indehiscent (1).<br />

50. Fruit lacking thickened sutures (0); fruit with thickened sutures that <strong>of</strong>ten separate from the valves in dehiscence (1).<br />

51. Fruit a capsule (0); fruit a berry (1).<br />

52. Seed with vascular bundle in the raphe (0); seed lacking vascular bundle in the raphe (1).<br />

53. Seed lacking fringe formed from differentiated cells (0); seed with fringe due to expanded and/or elongated, bulging cells (1).<br />

54. Seed lacking tails (0); seed with tails, i.e., expanded portion <strong>of</strong> testa at hilum and chalazal ends (1).<br />

55. Seed lacking unilateral wing (0); seed with unilateral wing (resulting from a flattening <strong>of</strong> chalazal end <strong>of</strong> seed, with cells <strong>of</strong> the wing not<br />

differentiated from those <strong>of</strong> the body) (1).<br />

56. Testa one-layered (except sometimes in the raphe) (0); testa with several cell layers (1).<br />

57. Testa cells isodiametric (in surface view) (0); testa cells much elongated (1).<br />

58. Methyl salicylate lacking (0); methyl salicylate present (1).<br />

59. Basic chromosome number 11 (0); 12 (1); 18 (2). [unordered]<br />

60. Toxic diterpenes <strong>of</strong> andromedotoxin type absent (0); toxic diterpenes <strong>of</strong> andromedotoxin type present (1).


1294 AMERICAN JOURNAL OF BOTANY<br />

[Vol. 86<br />

TABLE 3. Character values for species used in morphology-based cladistic analysis. V condition variable; ? condition unknown or not<br />

applicable.<br />

Species<br />

Character<br />

1 2 3 4 5 6<br />

123456789012345678901234567890123456789012345678901234567890<br />

Enkianthus campanulatus 000000000000000000000000000000000000000000000000000000000000<br />

Prionotes cerinthoides 0100000001011100000000020000000?000211?0000010000001000000?0<br />

Sprengelia incarnata 0100010101001101?0?200020000000?101211?000001000000100001010<br />

Agarista populifolia 1201001V010010000000V10101000000100011?00101100000010000101?<br />

Agarista salicifolia 10010011010010100000110101000000100011?0010110000001000010??<br />

Pieris floribunda 1000001001000000001011110110100100001010010110000001000010?1<br />

Pieris formosa 100000000100100000101111011010010000101001011000000100001011<br />

Pieris phillyreifolia 1000000001001000001011110110100110021010010110000001000000?1<br />

Pieris nana 11001001010010000010111101100001000110100V0110000001000010?1<br />

Craibiodendron yunnanense 11000001011110000010010101000000100111?0010110000001001010??<br />

Lyonia ferruginea 1100000101101000011011010100001010011V2001111010010101001011<br />

Lyonia lucida 110000010110100000101101011000101001102001111010010100001011<br />

Lyonia ligustrina 101000000010V00010101101011000101000102001111010010100001011<br />

Lyonia ovalifolia 111000010V10V000101011010110001010001020011110V0010100001011<br />

Vaccinium macrocarpon 01000001010000000001?00101000000000011?010011100102100000010<br />

Vaccinium meridionale 01000000010000100001110101000000000101010011101102100000010<br />

Satyria warszewiczii 01000001011000000001110101000000010011?01001110010210000V0?0<br />

Leucothoë fontanesiana 020000000100000000011111010000000001100001011000000110000001<br />

Leucothoë racemosa 010000000000000000011111010000000002100001011000000100000001<br />

Andromeda polifolia 0001000101001001?0?011010100010100001000000011000000100010011<br />

Zenobia pulverulenta 100000000000000000001101010000000001100101011000000100000001<br />

Chamaedaphne calyculata 0100001001000000020V11010100000V0002100010011000000100000001<br />

Diplycosia acuminata 01000000010100100001V10111010000V00211?01001100000010000112?<br />

Gaultheria miqueliana 000000000100000000011101010100000001100101011000000100000101<br />

Gaultheria shallon 0000010000100000000011101010101100000100101011000000100000001<br />

Pernettya tasmanica 00000000010000100001?102010101000000100100011000101100000001<br />

Tepuia cardonae 01000001110000100001100111000V0?000011?0100110001011000001??<br />

Oxydendrum arboreum 0200000V000000000001100100000000000011?001011000000100001011<br />

Note: 25: Enkianthus scored as ‘‘0’’ because it has free bracts, which are homologous to bracteoles. 28: the coding <strong>of</strong> this character for members<br />

<strong>of</strong> the Vaccinieae, e.g., Vaccinium and Satyria, is problematic; they are coded as ‘‘0’’ based on the assumption that the colorful, fleshy sepals in<br />

Gaultheria and Diplycosia developed independently from the fleshy fruits <strong>of</strong> Vaccinieae; developmental studies would be useful. 60: See also<br />

Hegnauer (1966) and Anderberg (1993).<br />

hairs (no. 47), and capsule with thickened sutures (no.<br />

50). Paired appendages on the filament (no. 39-2) are an<br />

additional synapomorphy under DELTRAN optimization.<br />

Agarista is supported by the apomorphies <strong>of</strong> leaves revolute<br />

in the bud (no. 4), leaves with a dense vein reticulum<br />

(no. 7), and stamens lacking appendages (no. 38).<br />

Gaultheria is monophyletic if Pernettya (as represented<br />

by P. tasmanica) is included, based on their fleshy calyx<br />

lobes (no. 28).<br />

The Lyonia group is monophyletic in some trees, but<br />

is paraphyletic in others because Andromeda polifolia is<br />

placed within the group (as sister to Agarista). Characters<br />

supportive <strong>of</strong> the monophyly <strong>of</strong> the Lyonia group include<br />

bands <strong>of</strong> fibers in the phloem (no. 1), S-shaped filaments<br />

(no. 33), disintegration tissue on back <strong>of</strong> anthers (no. 42),<br />

and more or less elongated testa cells (no. 57). Members<br />

<strong>of</strong> this group also have leaves with lignified epidermal<br />

cells (no. 13) and anomocytic stomata (no. 20-0). Both<br />

features also occur in Andromeda and in some trees function<br />

as synapomorphies. Generic relationships within the<br />

Lyonia group are poorly resolved, but Craibiodendron is<br />

always sister to Lyonia (Figs. 1, 2), a relationship consistently<br />

supported by their homogeneous pith (no. 2-1)<br />

and bifacial leaf midrib bundles (no. 11).<br />

The Gaultheria group is never monophyletic, although<br />

some <strong>of</strong> these genera, i.e., Gaultheria, Pernettya, Zenobia,<br />

Leucothoë, and also Chamaedaphne, are linked in a<br />

few cladograms by the base chromosome number <strong>of</strong> 11<br />

(no. 59). Gaultheria, Pernettya, and Zenobia <strong>of</strong>ten form<br />

a clade based on the presence <strong>of</strong> forked anther appendages<br />

(no. 40). The monophyly <strong>of</strong> Leucothoë is equivocal,<br />

but L. racemosa and L. fontanesiana are sometimes<br />

linked by their inflorescences that are exposed during the<br />

winter (no. 23), a feature that also evolved in Pieris (Fig.<br />

2).<br />

Finally, although not a major focus in this investigation,<br />

we note that Sprengelia incarnata and Prionotes<br />

cerinthoides form a clade in all trees that is supported by<br />

the lignified leaf epidermis (character no. 13), elongated<br />

epidermal cells (no. 14), pedicel with several bracteoles<br />

(no. 24-2), and smooth staminal filaments (no. 36-2).<br />

These two species represent the epacrid clade (see Crayn<br />

et al., 1996; Powell et al., 1996; Kron, 1997; Kron et al.,<br />

1998).<br />

Molecular data—Analysis <strong>of</strong> the rbcL data resulted in<br />

42 trees (L 503, CI 0.58, RI 0.50). The strict<br />

consensus (Fig. 3) indicates that few clades are well supported<br />

including: Vaccinieae (parjack 94), Lyonia (parjack<br />

99), and the Pieris floribunda P. formosa clade<br />

(parjack 90). Most <strong>of</strong> the structure in this tree collapses<br />

in trees one step longer. By contrast, the result <strong>of</strong> the<br />

matK analysis is much better resolved, especially within<br />

the Lyonia group (Fig. 4). In the matK analysis (four trees<br />

obtained, L 980, CI 0.65, RI 0.62) Vaccinieae<br />

and members <strong>of</strong> the Gaultheria group form a clade


September 1999] KRON ET AL.—PHYLOGENETICS OF ANDROMEDEAE<br />

1295<br />

Fig. 1. Strict consensus <strong>of</strong> all most parsimonious trees (36) from<br />

analysis <strong>of</strong> morphological data. Length (L) 139, Consistency Index<br />

(CI) 0.48, Retention Index (RI) 0.68. Decay values are below lines.<br />

Taxa traditionally placed in the tribe Andromedeae are in boldface.<br />

strongly supported by the data. Within this clade Vaccinieae<br />

are monophyletic and form a polytomy with three<br />

other clades. The sister relationship <strong>of</strong> Andromeda polifolia<br />

and Zenobia pulverulenta is strongly supported by<br />

the matK data and there is strong support for Chamaedaphne<br />

calyculata sister to Leucothoë racemosa. These<br />

results are different from the morphological analysis (Fig.<br />

1) where Andromeda is indicated as more closely related<br />

to the Lyonia group than the Gaultheria group. Similar<br />

to the morphological analysis, Lyonia and Pieris are<br />

monophyletic in the matK analysis. The Lyonia group<br />

(Lyonia, Craibiodendron, Pieris, Agarista) is monophyletic<br />

but not very strongly supported. In both the rbcL<br />

and matK analyses the Gaultheria group is placed in the<br />

same clade with a monophyletic Vaccinieae. In the matK<br />

analysis this large clade is strongly supported. The results<br />

<strong>of</strong> the rbcL analysis indicate only weak support for a<br />

Gaultheria group Vaccinieae clade, with the clade collapsing<br />

in trees one step longer than most parsimonious<br />

(Fig. 3). The Gaultheria group is not supported as monophyletic<br />

by either the rbcL or the matK analyses. Both<br />

analyses indicate a core Gaultheria clade that includes<br />

Tepuia, Diplycosia, Gaultheria, and Pernettya. In the<br />

rbcL analysis Diplycosia is sister to Tepuia, whereas in<br />

the matK analysis Diplycosia is included in a polytomy<br />

containing Gaultheria and Pernettya. However, none <strong>of</strong><br />

these relationships is strongly supported in either analysis.<br />

Combined analysis <strong>of</strong> rbcL and matK data resulted in<br />

four most parsimonious trees (L 1495, CI 0.92, RI<br />

0.58). The strict consensus (Fig. 5) is highly resolved<br />

with Oxydendrum sister to the remaining ingroup taxa.<br />

As in the separate molecular data analyses Agarista,<br />

Lyonia, and Pieris are monophyletic, but in the combined<br />

analysis they each have higher levels <strong>of</strong> support than in<br />

each individual analysis. In the combined rbcL and matK<br />

analysis Vaccinieae are monophyletic and unresolved<br />

with respect to the Andromeda Zenobia clade and the<br />

weakly supported clade that includes Leucothoë, Chamaedaphne,<br />

Tepuia, Diplycosia, Gaultheria, and Pernettya.<br />

In this analysis Gaultheria is paraphyletic with respect<br />

to Pernettya, a relationship also found in the rbcL<br />

and morphological analyses.<br />

Combined molecular and morphological analysis—<br />

This analysis resulted in the generation <strong>of</strong> a single most<br />

parsimonious tree <strong>of</strong> 1565 steps, a consistency index (CI)<br />

<strong>of</strong> 0.61, and a retention index (RI) <strong>of</strong> 0.59 (Fig. 6). The<br />

position <strong>of</strong> Oxydendrum as sister to the remaining ingroup<br />

taxa is strongly supported.<br />

The monophyly <strong>of</strong> the Lyonia group (incl. Agarista,<br />

Craibiodendron, Lyonia, and Pieris) is strongly supported<br />

(d8, 99% bootstrap). Noteworthy morphological features<br />

supportive <strong>of</strong> the monophyly <strong>of</strong> this group include<br />

bands <strong>of</strong> fibers in the phloem (no. 1), leaves with lignified<br />

epidermal cells (no. 13), S-shaped filaments (no. 33), and<br />

more or less elongated testa cells (no. 57). Within this<br />

clade, Agarista likely is sister to Pieris, and Craibiodendron<br />

is weakly supported as the sister group <strong>of</strong> Lyonia.<br />

Agarista, Pieris, and Lyonia are all strongly supported<br />

as monophyletic (Fig. 6); the monophyly <strong>of</strong> Craibiodendron<br />

was not assessed. Cladogram topology within<br />

the Lyonia group was identical to that discovered in our<br />

previous analysis <strong>of</strong> this group (Kron and Judd, 1997),<br />

and that paper should be consulted for a more detailed<br />

discussion <strong>of</strong> these generic and infrageneric relationships.<br />

Vaccinieae form a well-supported clade (d24, 100%<br />

bootstrap) with members <strong>of</strong> the Gaultheria group, Andromeda,<br />

and Chamaedaphne. A significant morphological<br />

synapomorphy <strong>of</strong> this clade is paracytic stomata (no.<br />

20). The Vaccinieae are clearly monophlyetic (d19, 100%<br />

bootstrap) and are supported by the same morphological<br />

characters as in the morphology-based analyses, with the<br />

addition <strong>of</strong> the anther tubules (no. 41). Such tubules evidently<br />

have evolved independently in a few other taxa,<br />

e.g., Chamaedaphne and the Tepuia Diplycosia clade.<br />

The remaining taxa, mainly members <strong>of</strong> the Gaultheria<br />

group, form a very weakly supported clade (d1, 50%<br />

bootstrap), which possibly is supported by the synapomorphy<br />

<strong>of</strong> a base chromosome number <strong>of</strong> 11 (although<br />

Andromeda has x 12). In addition, forked appendages<br />

(no. 40) occur on the anthers <strong>of</strong> Zenobia, Gaultheria, and<br />

Pernettya and may have been lost from the remaining<br />

taxa. Within this group, Andromeda and Zenobia are sister<br />

to a potential clade (d1, 63% bootstrap) containing<br />

Chamaedaphne, Leucothoë, Tepuia, Diplycosia, Gaultheria,<br />

and Pernettya. However, a core element, containing<br />

Diplycosia, Tepuia, Gaultheria, and Pernettya is well<br />

supported (d11, 100% bootstrap). Members <strong>of</strong> this clade<br />

are united by the presence <strong>of</strong> methyl salicylate (lost in<br />

some species <strong>of</strong> Gaultheria and Diplycosia). They may<br />

also share the apomorphy <strong>of</strong> sepals that are fleshy and<br />

colorful, but the latter feature is absent in Tepuia (and in<br />

some species <strong>of</strong> Pernettya) and may have evolved inde-


1296 AMERICAN JOURNAL OF BOTANY<br />

[Vol. 86<br />

Fig. 2. Representative most parsimonious tree (one <strong>of</strong> 36) from the analysis <strong>of</strong> morphological data. Phylogenetically informative characters are<br />

mapped onto the tree using ACCTRAN optimization (DELTRAN optimizations are discussed in the text). Solid bars—unique; hatched bars—occur<br />

more than once in the tree; open bars—reversals. Taxa traditionally placed in the tribe Andromedeae are in boldface.


September 1999] KRON ET AL.—PHYLOGENETICS OF ANDROMEDEAE<br />

1297<br />

Fig. 3. Strict consensus <strong>of</strong> 42 trees found in the parsimony analysis<br />

<strong>of</strong> rbcL data. L 503, CI 0.58, RI 0.50. Decay values are below<br />

lines, and parsimony jackknife values are above lines. Taxa traditionally<br />

placed in the tribe Andromedeae are in boldface.<br />

Fig. 4. Strict consensus <strong>of</strong> four trees found in the parsimony analysis<br />

<strong>of</strong> matK data. L 980, CI 0.65, RI 0.62. Decay values are<br />

below lines, and parsimony jacknife values are above lines. Taxa traditionally<br />

placed in the tribe Andromedeae are in boldface.<br />

Fig. 5. Strict consensus <strong>of</strong> four most parsimonious trees obtained<br />

from the combined analysis <strong>of</strong> rbcL and matK data. L 1495, CI <br />

0.62, RI 0.58. Decay values are below lines, and parsimony jacknife<br />

values are above lines. Taxa traditionally placed in the tribe Andromedeae<br />

in boldface.<br />

pendently in Diplycosia and Gaultheria. Tepuia and Diplycosia<br />

form a moderately well-supported clade (d3,<br />

71% bootstrap) supported by apical connate bracteoles<br />

(#25) and anther tubules (no. 41), as do Gaultheria and<br />

Pernettya (d4, 84% bootstrap), which may be supported<br />

by fleshy, colorful sepals. Gaultheria likely is paraphyletic,<br />

justifying the inclusion <strong>of</strong> Pernettya. The monophyly<br />

<strong>of</strong> Leucothoë is not at all supported.<br />

DISCUSSION<br />

Results <strong>of</strong> phylogenetic analyses based on morphology,<br />

rbcL, and matK sequences (Figs. 1–4) are generally<br />

similar, although analyses based on single character sets<br />

are less well resolved than the combined analyses (Figs.<br />

5–6). We note that the cladogram resulting from the analysis<br />

based on all three data sets, i.e., morphology and<br />

rbcL and matK sequences (Fig. 6), is better resolved and<br />

shows stronger statistical support for the various internal<br />

clades than any other tree produced in our analyses. This<br />

tree strongly supports the monophyly <strong>of</strong> the Lyonia group<br />

(including Lyonia, Craibiodendron, Agarista, and Pieris)<br />

and the Vaccinieae (including numerous inferior-ovaried<br />

genera), but provides only weak support for the Gaultheria<br />

group (circumscribed so as to include Andromeda and<br />

Chamaedaphne, as well as Zenobia, Leucothoë, Diplycosia,<br />

Tepuia, Gaultheria, and Pernettya). Recognition <strong>of</strong>


1298 AMERICAN JOURNAL OF BOTANY<br />

[Vol. 86<br />

Fig. 6. Single most parsimonious tree obtained from the combined<br />

analysis <strong>of</strong> morphology, rbcL, and matK data. L 1565, CI 0.61,<br />

RI 0.59. Decay values are below lines, and bootstrap values are above<br />

lines. Taxa traditionally placed in the tribe Andromedeae are in boldface.<br />

the Andromedeae, as traditionally circumscribed (see Stevens,<br />

1969, 1971, 1995), cannot be maintained, as genera<br />

<strong>of</strong> the Gaultheria group are more closely related to the<br />

Vaccinieae than they are to those <strong>of</strong> the Lyonia group.<br />

The strong support for the Lyonia group in the combined<br />

analysis indicates that the group should be given formal<br />

recognition, and it is here proposed as the tribe Lyonieae—Lyonieae<br />

Kron & Judd, tribus nov. Epidermis foliorum<br />

lignea; filamenta floris plerumque S-formis; testa<br />

cellulis plerumque elongatis. Type genus: Lyonia Nutt.<br />

Other nomenclatural proposals relating to these genera<br />

will be presented in connection with a new classification<br />

<strong>of</strong> the Ericaceae (Kron et al., unpublished data).<br />

The monophyly <strong>of</strong> the Gaultheria group is not supported<br />

in most <strong>of</strong> the analyses and only weakly supported<br />

in the molecular morphology analysis. In the combined<br />

rbcL and matK analysis the Gaultheria group is split into<br />

two clades that are not resolved with respect to Vaccinieae.<br />

Inclusion <strong>of</strong> the morphological data results in only<br />

a weakly supported Gaultheria group. This group needs<br />

further study, especially a more comprehensive sampling<br />

<strong>of</strong> Gaultheria, Pernettya, and Diplycosia. In addition,<br />

Leucothoë appears as paraphyletic or polyphyletic in<br />

most <strong>of</strong> the trees. Exceptions are in some <strong>of</strong> the morphological<br />

trees and in the rbcL analysis. In both cases the<br />

relationship <strong>of</strong> Leucothoë racemosa to L. fontanesiana is<br />

weakly supported.<br />

Andromeda, Chamaedaphne, and Oxydendrum <strong>of</strong>ten<br />

have been considered taxonomically isolated (Palser,<br />

1951, 1952; Stevens, 1969, 1971), but our results support<br />

an isolated position for only Oxydendrum (Fig. 6).<br />

Andromeda has several distinctive features, which<br />

probably represent autapomorphies, e.g., seeds with a<br />

strongly multilayered testa, campylotropous ovules (Palser,<br />

1952), the lack <strong>of</strong> calyx and corolla stomata, and lack<br />

<strong>of</strong> multicellular hairs. The linking <strong>of</strong> this genus with Zenobia<br />

was not anticipated and is not reinforced by any<br />

morphological characters included in our analyses. The<br />

base chromosome number <strong>of</strong> Andromeda, i.e., 12, also is<br />

aberrant given its placement with genera <strong>of</strong> the Gaultheria<br />

group, which are mainly x 11. However, it may be<br />

significant that Zenobia nearly lacks multicellular hairs,<br />

i.e., they are only associated with the leaf serrations, and<br />

if these were lost (as presumably occurred in the ancestors<br />

<strong>of</strong> Andromeda) then this genus, like Andromeda,<br />

would lack such hairs.<br />

Chamaedaphne is distinctive in its floral anatomy, embryology,<br />

campylotropous ovules (Palser, 1951, 1952),<br />

anthers with tubules, and distinctive lepidote indumentum<br />

(composed <strong>of</strong> peltate scales that are different in form<br />

from those <strong>of</strong> Lyonia sect. Lyonia; see Judd, 1979). Both<br />

molecular and morphological analyses place Chamaedaphne<br />

with members <strong>of</strong> the Gaultheria group. It is<br />

placed sister to Leucothoë racemosa in our combined<br />

analysis. Both have stamens with smooth filaments (no.<br />

36-2; see Fig. 4, where feature is homoplasious). Stevens<br />

(1969) noted that ‘‘in stomata, bracteole, seed and general<br />

anatomical characters it shows some similarity to the<br />

Gaultheria group.’’ For example, its stomata are positionally<br />

paracytic and testa cells isodiametric.<br />

Oxydendrum also is phenetically distinctive, mainly<br />

because <strong>of</strong> its floral anatomy (all traces to the floral organs<br />

leave an elongated axis separately; Palser, 1952),<br />

deeply hooked leaf midrib vascular bundle (Stevens,<br />

1969, 1971), anthers with terminal tubules, and terminal<br />

inflorescence that fruits in the same year in which the<br />

shoot on which it is borne was initiated (Lems, 1962).<br />

Cox (1948) considered Oxydendrum in its own tribe because<br />

its wood has a distinctive medullary ray type and<br />

a high percentage <strong>of</strong> porous vessel perforation plates. Our<br />

analyses suggest that these features represent a combination<br />

<strong>of</strong> autapomorphies and retained plesiomorphies.<br />

Our results are in general agreement with those <strong>of</strong> Stevens<br />

(1969, 1971), Judd (1979), and Middleton (1991) in<br />

supporting the recognition <strong>of</strong> two major clades within the<br />

genera traditionally placed within Andromedeae. Clearly,<br />

they also support the hypothesis (see Stevens, 1995) that<br />

the Andromedeae are not monophyletic, since Vaccinieae<br />

are closely related to members <strong>of</strong> the Gaultheria group.<br />

The close relationship <strong>of</strong> Vaccinieae to some genera traditionally<br />

placed within Andromedeae was first noted by<br />

Stevens (1969, 1971), and their phenetic similarity was<br />

evident to Watson, Williams, and Lance (1967).<br />

Higher level phylogenetic analyses <strong>of</strong> Ericaceae based<br />

on both nuclear and chloroplast DNA (see Kron and<br />

Chase, 1993; Kron, 1996, 1997; Kron et al., 1998) support<br />

the existence <strong>of</strong> a clade comprising Oxydendrum, the<br />

Gaultheria group, the Lyonia group, and the Vaccinieae,<br />

i.e., the ingroup taxa <strong>of</strong> the analyses herein presented.<br />

This monophyletic group, which in large part corresponds<br />

to the Vaccinioideae <strong>of</strong> Stevens (1969), probably should<br />

be given subfamilial rank within an expanded Ericaceae


September 1999] KRON ET AL.—PHYLOGENETICS OF ANDROMEDEAE<br />

1299<br />

(see Anderberg, 1993; Judd and Kron, 1993; Kron and<br />

Chase, 1993). The genera comprising this subfamily represent<br />

four clades, here listed in presumed order <strong>of</strong> divergence:<br />

(1) Oxydendrum, (2) Lyonia group, (3) Gaultheria<br />

group (as here recircumscribed), and (4) Vaccinieae.<br />

LITERATURE CITED<br />

ANDERBERG, A. A. 1993. Cladistic interrelationships and major clades<br />

<strong>of</strong> the Ericales. Plant Systematics and Evolution 184: 207–231.<br />

———. 1994. Cladistic analysis <strong>of</strong> Enkianthus with notes on the early<br />

diversification <strong>of</strong> the Ericaceae. Nordic <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 14: 385–<br />

401.<br />

ARTOPOEUS, A. 1903. Über den Bau und die Öffnungsweise der Antheren<br />

und Entwicklung der Samen der Erikaceen. Flora 92: 309–<br />

345.<br />

BAAS, P. 1985. Comparative leaf anatomy <strong>of</strong> Pernettya Gaud. (Ericaceae).<br />

Botanische Jahrbücher 105: 481–495.<br />

CHASE, M.W.,AND H. G. HILLS. 1991. Silica gel: an ideal material for<br />

field preservation <strong>of</strong> leaf samples for DNA studies. Taxon 40: 215–<br />

220.<br />

CHOU, Y.-L. 1952. Floral morphology <strong>of</strong> three species <strong>of</strong> Gaultheria.<br />

Botanical Gazette 114: 198–221.<br />

COPELAND, H. F. 1954. Observations on certain Epacridaceae. <strong>American</strong><br />

<strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 41: 215–222.<br />

COX, H. T. 1948. Studies in the comparative anatomy <strong>of</strong> the Ericales.<br />

II. Ericaceae—subfamily Arbutoideae. <strong>American</strong> Midland Naturalist<br />

40: 493–516.<br />

CRAYN, D. M., K. A. KRON, P.A.GADEK, AND C. J. QUINN. 1996.<br />

Delimitation <strong>of</strong> Epacridaceae: preliminary molecular evidence. Annals<br />

<strong>of</strong> <strong>Botany</strong> 77: 317–321.<br />

DORR, L. J. 1980. The reproductive biology and pollination ecology <strong>of</strong><br />

Zenobia (Ericaceae). Master’s thesis, University <strong>of</strong> North Carolina,<br />

Chapel Hill, NC.<br />

DOYLE, J., AND J. DOYLE. 1987. A rapid DNA isolation procedure for<br />

small quanities <strong>of</strong> fresh leaf tissue. Phytochemical Bulletin 19: 11–<br />

15.<br />

ERIKSSON, T., AND N. WIKSTRÖM. 1995. AUTODECAY, version 3.0.<br />

Stockholm, Sweden.<br />

FARRIS, J. S. 1988. Hennig86 reference, version 1.5. Port Jefferson<br />

Station, NY.<br />

———, V. A. ALBERT, M.KÄLLERSJO, D.LIPSCOMB, AND A. KLUGE.<br />

1996. Parsimony jackknifing outperforms neighbor-joining. Cladistics<br />

12: 99–124.<br />

HARBORNE, J. B. 1969. Gossypetin and herbacetin as taxonomic markers<br />

in higher plants. Phytochemistry 8: 177–183.<br />

———, AND C. A. WILLIAMS. 1973. A chemotaxonomic survey <strong>of</strong><br />

flavonoids and simple phenols in leaves <strong>of</strong> the Ericaceae. Botanical<br />

<strong>Journal</strong> <strong>of</strong> the Linnean Society 66: 37–54.<br />

HEGNAUER, R. 1966. Chemotaxonomie der Pflanzen 4. Daphniphyllaceae-Lythraceae.<br />

Birkhäuser, Basel.<br />

HERSEY, R. E., AND S. P. VANDER KLOET. 1976. Taxonomy and distribution<br />

<strong>of</strong> Gaultheria in the Caribbean. Canadian <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong><br />

54: 2465–2472.<br />

JOHNSON, L.A., AND D. E. SOLTIS. 1994. matK DNA sequences and<br />

phylogenetic reconstruction in Saxifragaceae s.s. Systematic <strong>Botany</strong><br />

19: 143–156.<br />

———. 1995. Phylogenetic inference in Saxifragaceae sensu stricto<br />

and Gilia (Polemoniaceae) using matK sequences. Annals <strong>of</strong> the<br />

Missouri Botanical Garden 82: 149–175.<br />

JUDD, W. S. 1979. Generic relationships in the Andromedeae (Ericaceae).<br />

<strong>Journal</strong> <strong>of</strong> the Arnold Arboretum 60: 477–503.<br />

———. 1981. A monograph <strong>of</strong> Lyonia (Ericaceae). <strong>Journal</strong> <strong>of</strong> the Arnold<br />

Arboretum 62: 63–209, 315–436.<br />

———. 1982. A taxonomic revision <strong>of</strong> Pieris (Ericaceae). <strong>Journal</strong> <strong>of</strong><br />

the Arnold Arboretum 63: 103–144.<br />

———. 1984. A taxonomic revision <strong>of</strong> the <strong>American</strong> species <strong>of</strong> Agarista<br />

(Ericaceae). <strong>Journal</strong> <strong>of</strong> the Arnold Arboretum 65: 255–342.<br />

———. 1986. A taxonomic revision <strong>of</strong> Craibiodendron (Ericaceae).<br />

<strong>Journal</strong> <strong>of</strong> the Arnold Arboretum 67: 441–469.<br />

———. 1990. A new variety <strong>of</strong> Lyonia (Ericaceae) from Puerto Rico.<br />

<strong>Journal</strong> <strong>of</strong> the Arnold Arboretum 71: 129–133.<br />

———. 1995a. 13. Lyonia Nuttall. In J. L. Luteyn [ed.], Ericaceae,<br />

part II—the superior-ovaried genera, 222–294. Flora Neotropica<br />

Monograph 66. New York Botanical Garden, Bronx, NY.<br />

———. 1995b. 14. Agarista D. Don ex G. Don. In J. L. Luteyn [ed.],<br />

Ericaceae, part II—the superior-ovaried genera, 295–344. Flora<br />

Neotropica Monograph 66. New York Botanical Garden, Bronx,<br />

NY.<br />

———. 1995c. 15. Pieris D. Don. In J. L. Luteyn [ed.], Ericaceae part<br />

II—the superior-ovaried genera, 345–350. Flora Neotropica Monograph<br />

66. New York Botanical Garden, Bronx, NY.<br />

———, AND P. M. HERMANN. 1990. Circumscription <strong>of</strong> Agarista bolibiensis<br />

(Ericaceae). Sida 14: 263–266.<br />

———, AND K. A. KRON. 1993. Circumscription <strong>of</strong> Ericaceae (Ericales)<br />

as determined by preliminary cladistic analyses based on<br />

morphological, anatomical, and embryological features. Brittonia<br />

45: 99–114.<br />

KRON, K. A. 1996. Phylogenetic relationships <strong>of</strong> Empetraceae, Epacridaceae,<br />

Ericaceae, Monotropaceae, Pyrolaceae: evidence from nuclear<br />

ribosomal 18s sequence data. Annals <strong>of</strong> <strong>Botany</strong> 77: 293–303.<br />

———. 1997. Exploring alternative systems <strong>of</strong> classification. Aliso 15:<br />

105–112.<br />

———, AND M. W. CHASE. 1993. Systematics <strong>of</strong> the Ericaceae, Empetraceae,<br />

Epacridaceae, and related taxa based upon rbcL sequence<br />

data. Annals <strong>of</strong> the Missouri Botanical Garden 80: 735–<br />

741.<br />

———, AND W. S. JUDD. 1997. [1998.] Systematics <strong>of</strong> the Lyonia<br />

group (Andromedeae, Ericaceae) and the use <strong>of</strong> species as terminals<br />

in higher-level cladistic analyses. Systematic <strong>Botany</strong> 22: 479–<br />

492.<br />

———, AND J. M. KING. 1996. Cladistic relationships <strong>of</strong> Kalmia, Leiophyllum,<br />

and Loiseleuria (Phyllodoceae, Ericaceae) based on rbcL<br />

and nrITS data. Systematic <strong>Botany</strong> 21: 17–29.<br />

———, R. FULLER, D.M.CRAYN, P.A.GADEK, AND C. J. QUINN. 1999.<br />

Phylogenetic relationships <strong>of</strong> epacrids and vaccinioids (Ericaceae<br />

s.l.) based on matK sequence data. Plant Systematics and Evolution<br />

216: in press.<br />

LEINS, P. 1964. Entwicklungsgeschichtliche Studien an Ericales-Blüten.<br />

Botanische Jahrbücher 83: 57–88.<br />

LEMS, K. 1962. Adaptive radiation in the Ericaceae. I. Shoot development<br />

in the Andromedeae. Ecology 43: 524–528.<br />

———. 1964. Evolutionary studies in the Ericaceae. II. Leaf anatomy<br />

as a phylogenetic index in the Andromedeae. Botanical Gazette<br />

125: 178–186.<br />

LUTEYN, J. L. 1991. Key to the subfamilies and genera <strong>of</strong> neotropical<br />

Ericaceae. Nordic <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 11: 623–627.<br />

———. 1995a. 16. Tepuia. In J. L. Luteyn [ed.], Ericaceae, part II—<br />

the superior-ovaried genera, 351–364. Flora Neotropica Monograph<br />

66. New York Botanical Garden, Bronx, NY.<br />

———. 1995b. 17. Pernettya. In J. L. Luteyn [ed.], Ericaceae, part<br />

II—the superior-ovaried genera, 365–383. Flora Neotropica Monograph<br />

66. New York Botanical Garden, Bronx, NY.<br />

———. 1995c. 18. Gaultheria. In J. L. Luteyn [ed.], Ericaceae, part<br />

II—the superior-ovaried genera, 384–488. Flora Neotropica Monograph<br />

66. New York Botanical Garden, Bronx, NY.<br />

———. 1996. Ericaceae. In G. Harling and L. Andersson [eds.], Flora<br />

<strong>of</strong> Ecuador 54. Council for Nordic Publications in <strong>Botany</strong>, Copenhagen.<br />

———, W. S. JUDD, S.P.VANDER KLOET, L.J.DORR, G.D.WALLACE,<br />

K. A. KRON, P.F.STEVENS, AND S. E. CLEMANTS. 1996. Ericaceae<br />

<strong>of</strong> the southeastern United States. Castanea 61: 101–144.<br />

MATTHEWS, J. R., AND E. M. KNOX. 1926. The comparative morphology<br />

<strong>of</strong> the stamen in the Ericaceae. Transactions and Proceedings<br />

<strong>of</strong> the Botanical Society <strong>of</strong> Edinburgh 29: 243–281.<br />

MELVIN, N. C. 1980. A systematic investigation <strong>of</strong> the genus Leucothoë<br />

(Ericaceae). Ph.D. dissertation, Miami University, Oxford, OH.<br />

MIDDLETON, D. J. 1991a. Taxonomic studies in the Gaultheria group<br />

<strong>of</strong> genera <strong>of</strong> the tribe Andromedeae (Ericaceae). Edinburgh <strong>Journal</strong><br />

<strong>of</strong> <strong>Botany</strong> 48: 283–306.<br />

———. 1991b. Infrageneric classification <strong>of</strong> the genus Gaultheria L.<br />

(Ericaceae). Botanical <strong>Journal</strong> <strong>of</strong> the Linnean Society 106: 229–<br />

258.


1300 AMERICAN JOURNAL OF BOTANY<br />

[Vol. 86<br />

———, AND C. C. WILCOCK. 1990a. A critical examination <strong>of</strong> the<br />

status <strong>of</strong> Pernettya as a genus distinct from Gaultheria. Edinburgh<br />

<strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 47: 291–301.<br />

———, AND ———. 1990b. Chromosome counts in the genus Gaultheria<br />

and related genera. Edinburgh <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 47: 303–<br />

313.<br />

MISHLER, B. D., M. J. DONOGHUE, AND V. A. ALBERT. 1991. The decay<br />

index as a measure <strong>of</strong> relative robustness within a cladogram. 10th<br />

Willi Hennig Society Meeting, Toronto (Abstract).<br />

NIEDENZU, F. 1890. Über den anatomischen Bau der Laubblätter der<br />

Arbutoideae und Vaccinioideae in Beziehung zu ihrer systematischen<br />

Gruppierung und geographischen Verbreitung. Botanische Jahrbücher<br />

11: 134–263.<br />

ODELL, E. A., AND S. P. VANDER KLOET. 1991. The utility <strong>of</strong> stem<br />

characters in the classification <strong>of</strong> Vaccinium L. (Ericaceae). Taxon<br />

40: 273–283.<br />

OLMSTEAD, R. G., H. J. MICHAELS, K.M.SCOTT, AND J. D. PALMER.<br />

1992. Monophyly <strong>of</strong> the Asteridae and identification <strong>of</strong> their major<br />

lineages inferred from DNA sequences <strong>of</strong> rbcL. Annals <strong>of</strong> the Missouri<br />

Botanical Garden 79: 249–265.<br />

PALSER, B. F. 1951. Studies <strong>of</strong> floral morphology in the Ericales. I.<br />

Organography and vascular anatomy in the Andromedeae. Botanical<br />

Gazette 112: 447–485.<br />

———. 1952. Studies <strong>of</strong> floral morphology in the Ericales. II. Megagametophyte<br />

development in the Andromedeae. Botanical Gazette<br />

114: 33–52.<br />

———. 1954. Studies <strong>of</strong> floral morphology in Ericales. III. Organography<br />

and vascular anatomy <strong>of</strong> several species <strong>of</strong> Arbuteae. Phytomorphology<br />

4: 335–354.<br />

———. 1958. Studies <strong>of</strong> floral morphology in the Ericales. IV. Observations<br />

<strong>of</strong> three members <strong>of</strong> the Gaultherieae. Transactions <strong>of</strong> the<br />

Illinois Academy <strong>of</strong> Sciences 51: 24–34.<br />

———. 1961a. Studies <strong>of</strong> floral morphology in the Ericales. V. Organography<br />

and vascular anatomy in several United States species <strong>of</strong><br />

the Vacciniaceae. Botanical Gazette 123: 79–111.<br />

———. 1961b. Some aspects <strong>of</strong> embryology in the Ericales. Recent<br />

Advances in <strong>Botany</strong> 1: 685–689 (IX International Botanical Congress,<br />

Montreal, 1959). University <strong>of</strong> Toronto Press, Toronto.<br />

PATERSON, B. R. 1961. Studies <strong>of</strong> floral morphology in the Epacridaceae.<br />

Botanical Gazette 122: 259–279.<br />

POWELL, J. M., D. M. CRAYN, P.A.GADEK, C.J.QUINN, D.A.MOR-<br />

RISON, AND A. R. CHAPMAN. 1996. A re-assessment <strong>of</strong> relationships<br />

within Epacridaceae. Annals <strong>of</strong> <strong>Botany</strong> 77: 305–315.<br />

———, D. A. MORRISON, P.A.GADEK, D.M.CRAYN, AND C. J. QUINN.<br />

1997. Relationships and generic concepts within Styphelieae (Epacridaceae).<br />

Australian Systematic <strong>Botany</strong> 10: 15–29.<br />

RAO, T.A., AND S. CHAKRABORTI. 1985. The veinlet syndrome in the<br />

tribe Andromedeae (Ericaceae). Proceedings <strong>of</strong> the Indian Academy<br />

<strong>of</strong> Sciences 94: 639–654.<br />

SAFIJOWSKA, L. D. 1960. Male gametophyte in Enkianthus. Botanical<br />

Gazette 121: 190–195.<br />

SAMUELSSON, G. 1913. Studien über die Entwicklungsgeschichte der<br />

Blüten einiger Bicornes-Types. Svensk Botanisk Tidskrift 7: 97–<br />

188.<br />

SLEUMER, H. 1959. Studien über die Gattung Leucothoë D. Don. Botanische<br />

Jahrbücher 78: 435–480.<br />

———. 1966. Ericaceae. Flora Malesiana series I, 6: 469–668.<br />

———. 1967. Ericaceae. Flora Malesiana series I, 6: 669–914.<br />

STEELE, K.P., AND R. VILGALYS. 1994. Phylogenetic analyses <strong>of</strong> Polemoniaceae<br />

using nucleotide sequences <strong>of</strong> the plastid gene matK.<br />

Systematic <strong>Botany</strong> 19: 126–142.<br />

STEVENS, P. F. 1969. Taxonomic studies in the Ericaceae. Ph.D. dissertation,<br />

University <strong>of</strong> Edinburgh, Edinburgh.<br />

———. 1970a. Agauria and Agarista: an example <strong>of</strong> tropical transatlantic<br />

affinity. Notes <strong>of</strong> the Royal Botanic Garden, Edinburgh 30:<br />

341–359.<br />

———. 1970b. Calluna, Cassiope and Harrimanella: a taxonomic and<br />

evolutionary problem. New Phytologist 69: 1131–1148.<br />

———. 1971. A classification <strong>of</strong> the Ericaceae: subfamilies and tribes.<br />

Botanical <strong>Journal</strong> <strong>of</strong> the Linnean Society 64: 1–53.<br />

———. 1991. Character states, morphological variation, and phylogenetic<br />

analysis: a review. Systematic <strong>Botany</strong> 16: 553–583.<br />

———. 1995. Familial and infrafamilial relationships, In J. L. Luteyn<br />

[ed.], Ericaceae, part II—the superior-ovaried genera, 1–12. Flora<br />

Neotropica Monograph 66. New York Botanical Garden, Bronx,<br />

NY.<br />

SWOFFORD, D. L. 1993. PAUP: Phylogenetic analysis using parsimony,<br />

version 3.1. Computer program distributed by the Illinois Natural<br />

History Survey, Champaign, IL.<br />

TOWERS, G. H. N., A. TSE, AND W. S. G. MAAS. 1966. Phenolic acids<br />

and phenolic glycosides <strong>of</strong> Gaultheria species. Phytochemistry 5:<br />

677–681.<br />

VANDER KLOET, S. P. 1983. The taxonomy <strong>of</strong> Vaccinium sect. Cyanococcus:<br />

a summation. Canadian <strong>Journal</strong> <strong>of</strong> <strong>Botany</strong> 61: 256–266.<br />

———. 1988. The genus Vaccinium in North America. Agriculture<br />

Canada Publication 1828.<br />

VILLAMIL, P.H. DE, AND B. F. PALSER. 1981. Studies <strong>of</strong> floral morphology<br />

in the Ericales. IX. Organography, vascular anatomy and<br />

magagametophyte in three species <strong>of</strong> Gaultherieae. Phytomorphology<br />

30: 250–265.<br />

WAGNER, W. H. 1980. Origin and philosophy <strong>of</strong> the groundplan-divergence<br />

method <strong>of</strong> cladistics. Systematic <strong>Botany</strong> 5: 173–193.<br />

WATSON, L. 1962. The taxonomic significance <strong>of</strong> stomatal distribution<br />

and morphology in Epacridaceae. New Phytologist 61: 36–40.<br />

———. 1965. The taxonomic significance <strong>of</strong> certain anatomical variations<br />

among Ericaceae. Botanical <strong>Journal</strong> <strong>of</strong> the Linnean Society<br />

59: 111–125.<br />

———, W. T. WILLIAMS, AND G. N. LANCE. 1967. A mixed-data approach<br />

to Angiosperm taxonomy: the classification <strong>of</strong> Ericales.<br />

Proceedings <strong>of</strong> the Linnean Society, London 178: 25–35.<br />

WEINS, J. J. 1998. The accuracy <strong>of</strong> methods for coding and sampling<br />

higher-level taxa for phylogenetic analysis: a simulation study. Systematic<br />

Biology 47: 397–413.<br />

WILEY, E. O., D. SIEGEL-CAUSEY, D.R.BROOKS, AND V. A. FUNK. 1991.<br />

The compleat cladist: a primer <strong>of</strong> phylogenetic procedures. University<br />

<strong>of</strong> Kansas Museum <strong>of</strong> Natural History Special Publication<br />

Number 19, Lawrence, KS.<br />

WOOD, C. E., JR. 1961. The genera <strong>of</strong> Ericaceae in the southeastern<br />

United States. <strong>Journal</strong> <strong>of</strong> the Arnold Arboretum 42: 10–80.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!